Current state of artificial intelligence in liver transplantation

Abstract

Over the past few decades, substantial progress has been made in the field of liver transplantation. Yet, challenges remain in the field due to an increasing organ allograft shortage as well as significant waitlist mortality. With these challenges, organ allocation policies have been developed and are constantly being modified to result in more efficient organ allocation. One tool that has been explored to improve the field of liver transplantation is artificial intelligence, which is an umbrella term for techniques such as machine learning and deep learning. This review article explores the use of artificial intelligence in the field of liver transplantation. Specifically, studies have shown potential applications of artificial intelligence in improving waitlist mortality models, assessing allograft characteristics, using large language models for research question development and patient education, developing post-transplant models, as well as predicting multiple risk factors such as cardiovascular disease, infection, graft failure, malignancy, graft fibrosis, and pneumonia. However, even with these studies, several limitations for the use of artificial intelligence exist such as biased data sets leading to biased model development, lack of extensive validation of the artificial intelligence models, and the need for large datasets for model development. With additional studies evaluating the use of artificial intelligence and wide-scale validation of these studies highlighted, the use of artificial intelligence may transform the field of transplantation in the future.